Topic Models • 2 main outputs of interest: 1. Topic Prevalence Matrix 2. Topic Content Matrix includes kosovo Ecolo EcologicalReductionism SocialReductionism Ecolo Text Analysis(Social Sciences) Text Mining(Computer Science) Topic 1 (e.g. Deforestation) Topic 2 (e.g. Conservation) Topic 3 (e.g. Political Ecology) Topic Models In the EU journalistic field, a high prevalence of issues around GMOs, health, and consumer labeling in the late 1990s has given way to more abstract and distant phenomena, such as deforestation in producing regions and climate change since themid-2000s. This has been accompanied with a broaderperspective of different stake-holders, but also with a strongregional focus on the Amazon biome. Less attention has been directed at immediate concerns for local communities, such as land conflicts or disputes over intellectual property rights, which are more prevalent in the advocacy field or the journalistic fields in Brazil and Argentina • Why do you think these findings may be important? Text Mining Environmental Communication - Tracing the framing of problems with textminingFinn Mempel - f.mempel@uab.cat
1
  1. Intro
  2. Title
  3. Index_1
  4. Env_Comm_1
  5. Env_Comm_2
  6. Env_Comm_3
  7. Env_Comm_4
  8. Env_Comm_5
  9. Env_Comm_6
  10. Index_2
  11. Text_M_1
  12. Text_M_2
  13. Text_M_3
  14. Text_M_4
  15. Text_M_5
  16. Text_M_6
  17. Text_M_6
  18. Index_3
  19. Soy_1
  20. Soy_2
  21. Soy_3
  22. Soy_4
  23. Soy_5
  24. Soy_6
  25. Soy_7
  26. Soy_8
  27. Soy_9
  28. Soy_10
  29. Soy_11
  30. Soy_12
  31. Soy_13
  32. Soy_14
  33. Soy_15
  34. Soy_16
  35. Soy_17
  36. Soy_18
  37. Soy_19
  38. Soy_20
  39. Soy_21
  40. Soy_22
  41. Soy_23
  42. Soy_24
  43. Index_4
  44. Questions
  45. Def_1
  46. Def_2
  47. Def_3